Skip to main content
BenchmarkComputer Visionv2017

ADE20K Segmentation

by Zhou et al. / MIT CSAIL · open-source · Last verified 2026-03-17

ADE20K is the benchmark for semantic scene parsing, containing 25,000 images densely annotated with 150 semantic categories. Mean Intersection over Union (mIoU) is the standard metric, and it drives progress in perception systems for autonomous driving, robotics, and scene understanding.

https://ade20k.csail.mit.edu
B+
B+Good
Adoption: AQuality: AFreshness: C+Citations: A+Engagement: F

Specifications

License
BSD 3-Clause
Pricing
open-source
Capabilities
evaluation, semantic-segmentation, scene-parsing
Integrations
Use Cases
model-evaluation, computer-vision, autonomous-driving
API Available
No
Evaluated Models
mask2former, segformer-b5, internimage-xl, swin-large
Metrics
miou, pixel-accuracy
Methodology
Single-scale evaluation on the 2,000-image validation set. mIoU computed by averaging per-class IoU across all 150 semantic categories. Pixel accuracy reported as a secondary metric.
Last Run
2025-11-20
Tags
semantic-segmentation, scene-parsing, vision, miou, dense-prediction
Added
2026-03-17
Completeness
100%

Index Score

76
Adoption
88
Quality
89
Freshness
58
Citations
92
Engagement
0

Explore the full AI ecosystem on Agents as a Service